Intelligent military physical fitness anti-dizziness training device and method

Abstract

This invention belongs to the field of military physical training technology. It provides an intelligent military physical fitness anti-vertigo training device and method, comprising: a three-degree-of-freedom electric rotating platform, a physiological information acquisition module, a VR immersive interaction module, a platform terminal control and AI central system, and a database storage module. The platform terminal control and AI central system includes: an equipment control unit, an AI assessment and intelligent intervention unit, an automatic scoring unit, and a communication interface unit. The AI ​​assessment and intelligent intervention unit includes a benchmark construction subunit, a real-time monitoring and analysis subunit, and a multi-level intervention triggering and execution subunit. The benchmark construction subunit acquires the trainee's basic information and historical training data. This invention uses a multi-target emergency priority analysis process to address simultaneous emergency scenarios involving multiple trainees. It assesses priority by combining intervention level classification and multi-level ranking, thereby ensuring that high-risk targets are handled first.

Description

Technical Field

[0001] This invention belongs to the field of military physical training technology, specifically an intelligent military physical anti-vertigo training device and method. Background Technology

[0002] Vertigo and spatial disorientation are significant issues affecting the combat physical fitness and safety of military personnel such as pilots, naval crew members, armored soldiers, and airborne troops. Therefore, anti-vertigo training is a crucial component of military physical training.

[0003] Currently, anti-vertigo training equipment commonly used by the military, such as fixed rotating chairs, spiral ladders, and fixed rollers, has significant limitations:

[0004] In terms of driving methods, the reliance on trainees' own exertion makes it difficult to precisely control and quantify rotational speed and acceleration, resulting in poor consistency in training intensity and hindering the implementation of scientifically progressive training. Regarding training modes, only simple vestibular stimulation is provided, lacking comprehensive simulation of complex environments related to combat, such as visual, auditory, and cognitive loads, leading to low integration of training and combat. In terms of assessment methods, reliance is placed on trainees' subjective feelings (such as gait observation after stopping and self-reporting) or simple timing and counting, lacking an objective and quantitative assessment system. In terms of safety, the inability to monitor trainees' physiological state in real time poses significant safety risks and makes it difficult to prevent accidents caused by overtraining. In terms of efficiency, training, assessment, and recording all require manual operation, resulting in low management efficiency and difficulty in meeting the needs of large-scale, personalized training.

[0005] Therefore, the present invention provides an intelligent military physical fitness anti-vertigo training device and method. Summary of the Invention

[0006] In order to overcome the shortcomings of the prior art, at least one technical problem raised in the background art is solved.

[0007] The technical solution adopted by the present invention to solve its technical problem is: an intelligent military physical fitness anti-vertigo training device, including: a three-degree-of-freedom electric rotating platform, a physiological information acquisition module, a VR immersive interaction module, a platform terminal control and AI central system, and a database storage module;

[0008] The platform terminal control and AI central system includes: a device control unit, an AI evaluation and intelligent intervention unit, an automatic scoring unit, and a communication interface unit;

[0009] The AI ​​assessment and intelligent intervention unit includes a benchmark construction subunit, a real-time monitoring and analysis subunit, and a multi-level intervention triggering and execution subunit.

[0010] Benchmark building sub-unit: Obtain the trainer's basic information and historical training data, and dynamically set the benchmark through a sliding window algorithm;

[0011] Real-time monitoring and analysis subunit: Acquires trainees' heart rate and blood oxygen saturation data in real time, performs heart rate variability analysis and blood oxygen trend analysis, and outputs comprehensive load index based on the analysis results;

[0012] Multi-level intervention triggering and execution sub-unit: Based on the comprehensive load index output, trigger tiered early warning and intervention execution.

[0013] An anti-vertigo training method includes the following steps:

[0014] Trainees log in through a facial recognition system, which automatically retrieves their personal profiles. Trainees then put on a physiological wristband, VR headset, headphones, and earpiece.

[0015] Instructors can select the standard training mode or customize the rotation parameters and question bank for this training session through the equipment control interface;

[0016] Training begins, the three-degree-of-freedom rotating platform starts moving, and the VR headset scene starts synchronously. While experiencing rotational stimulation, the trainee observes the situation in the VR headset environment and answers questions posed by the system.

[0017] Real-time collection and analysis of physiological information data; if any abnormality is detected, immediate intervention is performed according to the preset level.

[0018] After training, a detailed evaluation report is generated, and the intensity and focus of the next training session can be recommended based on historical data.

[0019] The beneficial effects of this invention are as follows:

[0020] This invention utilizes a three-degree-of-freedom electric rotary platform driven by an independent servo motor and controlled by a high-precision encoder. This allows for precise control of rotational speed, acceleration, and motion dimensions, enabling the scientific increase and standardization of training intensity.

[0021] This invention uses a VR immersive interactive module to simulate real combat scenarios such as tank bumps and warship swaying, and superimposes cognitive tasks such as tactical calculations and enemy-friend identification. It fills the gap of traditional equipment that only provides vestibular stimulation and is disconnected from training and combat, allowing trainees to enhance their combat skills while enduring dizziness stimulation.

[0022] This invention acquires heart rate and blood oxygen data in real time, combines AI assessment and intelligent intervention unit quantitative analysis to output objective comprehensive load indicators, and with a three-level intervention mechanism, effectively prevents safety hazards such as fainting and vestibular damage caused by overtraining, and greatly improves the safety and scientific nature of training.

[0023] This invention employs a multi-target emergency priority analysis process to assess the priority of high-risk targets in simultaneous emergency scenarios involving multiple trainees by combining intervention level classification with multi-level ranking. Attached Figure Description

[0024] The invention will now be further described with reference to the accompanying drawings.

[0025] Figure 1 This is a schematic diagram of an intelligent military physical fitness anti-vertigo training device.

[0026] Figure 2 This is a graphical interface diagram of the device control unit in this invention;

[0027] Figure 3 This is an architecture diagram of the AI ​​assessment and intelligent intervention unit in this invention;

[0028] Figure 4 This is a logical flowchart of the target emergency priority analysis process in this invention;

[0029] Figure 5 This is a step-by-step diagram of an intelligent military physical fitness anti-vertigo training method. Detailed Implementation

[0030] To make the technical means, creative features, objectives and effects of this invention easier to understand, the invention will be further described below in conjunction with specific embodiments.

[0031] Example 1

[0032] Please see Figure 1 and Figure 2 As shown in the embodiment of the present invention, an intelligent military physical fitness anti-vertigo training device includes:

[0033] The three-degree-of-freedom electric rotary platform, as the core mechanical structure, is made of high-strength steel to ensure stability and safety. It has three rotational degrees of freedom: pitch, roll, and yaw, each driven by an independent frequency-modulated servo motor. With the help of a high-precision encoder, the speed, direction, and acceleration can be precisely programmed and controlled. The platform integrates a voice receiver, which can respond to the trainee's voice commands. Emergency stop buttons are located on the control console and the handheld remote control to ensure safe control in emergency situations.

[0034] Physiological information acquisition module: Uses military-grade or high-precision civilian wristbands to collect trainees' physiological indicators in real time. The data is transmitted to the central processing system in real time via wireless means (such as Bluetooth 5.0).

[0035] Physiological indicators include at least: heart rate (HR) and blood oxygen saturation (SpO2).

[0036] The VR immersive interaction module includes: a VR headset unit, a display unit, and a headset and headphone unit;

[0037] VR headset unit: Provides an immersive visual experience, simulating specific military sports scenarios;

[0038] Specific military sports scenarios include, but are not limited to: tank jolting, warship rocking, helicopter flight, and parachute drop.

[0039] Display Unit: Overlays test questions in the VR scene. Test question types include, but are not limited to: tactical calculations (such as ballistic estimation), friend or foe identification (such as quickly judging the attributes of targets in the scene), equipment operation procedures, and battlefield decision-making.

[0040] Headset and headphone unit: The headset is used to collect the trainee's voice answers, and the headphone is used to play ambient sound effects and instructions. The built-in offline speech recognition engine converts speech into text and compares and scores the answers.

[0041] Platform terminal control and AI hub module: Adopting a modular software design, it can be deployed on a ruggedized PC or tablet, including: device control unit, AI assessment and intelligent intervention unit, automatic scoring unit and communication interface unit;

[0042] Equipment control unit: Provides a graphical interface that allows instructors to set training programs, including but not limited to: rotation axis, speed curve, number of forward and reverse rotations, and training duration.

[0043] AI assessment and intelligent intervention unit: Real-time analysis of heart rate variability and blood oxygen trends. Once it is determined that the trainee's physiological load is close to the preset safety threshold (the safety threshold can be adjusted according to military branch, age, and individual historical data), the system will sequentially initiate three levels of intervention.

[0044] Automatic scoring unit: After training, the system automatically generates a report. The scoring system integrates physiological indicators (such as average heart rate, blood oxygen saturation rate, and recovery speed) and cognitive task performance (answer accuracy and reaction time) to form vestibular stability score, physiological stress score, and cognitive efficacy score, providing instructors with a scientific and comprehensive evaluation basis.

[0045] Communication interface unit: Unifies the management of drivers, connection status detection and data communication for all hardware devices.

[0046] Database storage module: Used to store all training data, including personnel information, physiological data logs, training video recordings (optional), test question bank, evaluation reports, etc., supporting data traceability and big data analysis, and providing support for optimizing training programs.

[0047] Example 2

[0048] Based on the foregoing embodiments, please refer to Figure 2 As shown in the embodiment of the present invention, an intelligent military physical fitness anti-vertigo training device includes:

[0049] AI assessment and intelligent intervention unit: Real-time analysis of heart rate variability and blood oxygen trends. Once it is determined that the trainee's physiological load is close to the preset safety threshold (the safety threshold can be adjusted according to military branch, age, and individual historical data), the system will sequentially initiate three levels of intervention.

[0050] In some embodiments, the AI ​​assessment and intelligent intervention unit includes a benchmark construction subunit, a real-time monitoring and analysis subunit, and a multi-level intervention triggering and execution subunit.

[0051] Benchmark building sub-unit: Obtain the trainer's basic information and historical training data, and dynamically set the benchmark through a sliding window algorithm;

[0052] In this sub-unit, the trainee is identified and logged into the system based on facial recognition.

[0053] Retrieve trainees' basic profile information, including fixed attributes and health labels, to clarify initial individual differences;

[0054] Among them, the fixed attributes include at least: troop type, age, height and weight;

[0055] Health indicators include whether there are underlying conditions such as sinus arrhythmia or previous vestibular dysfunction.

[0056] Identify whether the trainee has historical training data;

[0057] If the trainee has historical training data, extract the physiological indicators and corresponding training plans from their past 3-5 anti-vertigo training sessions to form a historical data pool, and extract the initial physiological data that is closest to the current state in the historical training as the initial baseline.

[0058] Initial baselines include: resting heart rate reference value, baseline blood oxygen saturation reference value, and baseline heart rate variability (HRV) reference value;

[0059] For example, if a trainee participated in training 3 months ago, with a resting heart rate of 72 bpm and blood oxygen 97.5%, and there are no recent changes in underlying diseases or significant fluctuations in physical fitness, then "72 bpm (heart rate), 97.5% (blood oxygen), 48 ms (HRV)" will be used as the initial benchmark. If there are multiple resting records in the historical data, then "the most recent resting data without training fatigue" will be used (such as data with a training interval of more than 1 week and no interfering factors such as muscle soreness or lack of sleep) to ensure that the initial benchmark is close to the current physiological state.

[0060] If the trainee is training for the first time and has no historical training data, guide them to complete a 5-minute resting state monitoring, wear a military-grade wristband, remain seated, and collect the resting heart rate, baseline blood oxygen saturation, and baseline heart rate variability (such as RR interval standard deviation SDNN) during this period as the initial baseline.

[0061] Preprocessing was performed on the collected historical data and resting data to remove outliers;

[0062] The preprocessed physiological indicators under different training scenarios (such as heart rate data under low-intensity rotation vs. high-intensity rotation) are converted into a unified dimension (such as the percentage of fluctuation relative to the resting value).

[0063] The significance of this conversion lies in the fact that directly comparing the original heart rate only reflects numerical differences, but cannot eliminate the interference of different scene intensities (high-intensity scenes themselves consume more physical strength, so the absolute value of the heart rate is naturally higher); while converting it to a relative fluctuation percentage can uniformly reflect the degree of deviation of the trainee's physiological load in the current scene from their resting state, and can more objectively reflect that the load stimulation of high-intensity scenes is significantly higher for the trainee, avoiding the benchmark deviation caused by the different intensity of the scene itself.

[0064] For example, assume that trainee A's resting heart rate reference value is 70 beats / min (bpm), which will serve as the reference benchmark for all subsequent scenario data conversions;

[0065] Scenario 1: Low-intensity rotation training (5 revolutions per minute, no cognitive task);

[0066] Heart rate data collected in real time during training: average 84 bpm;

[0067] Converted to a percentage fluctuation relative to the resting value: Percentage fluctuation = [(Real-time value - Resting value) / Resting value] × 100% = [(84 - 70) / 70] × 100% = 20%;

[0068] Scenario 2: High-intensity rotation training (15 revolutions per minute, superimposed with tactical calculation tasks);

[0069] Heart rate data collected in real time during training: average 105 bpm;

[0070] Convert to a percentage of fluctuation relative to the resting value: Percentage of fluctuation = [(105-70) / 70] × 100% = 50%.

[0071] Using a 10-minute training duration as the sliding window unit, and combining the trainee's data from the first three low-intensity adaptive training sessions (such as pure vestibular stimulation training with a rotation speed ≤ 8 rpm and no cognitive task), the initial baseline was iteratively calibrated.

[0072] For example: If a trainee's initial resting heart rate is 70 bpm, and the resting heart rate fluctuation during the first low-intensity training is ≤5% (67-73 bpm), then "70±3 bpm" is set as the temporary heart rate baseline; if the resting heart rate stabilizes at 68 bpm during the second training session, and the maximum tolerated heart rate during high-intensity training increases by 8% compared to the first time, then the heart rate baseline is updated to "68±4 bpm" using a sliding window algorithm, while simultaneously adjusting the blood oxygen level (e.g., from 98% to 97.5%).

[0073] Finally, the baseline was adjusted based on the characteristics of different troop types;

[0074] For example, the safe upper limit of heart rate for pilots is set at 1.9 times the resting heart rate reference value because the vestibular load is higher in actual combat, while that for ordinary infantrymen is set at 1.8 times the resting heart rate reference value.

[0075] For trainees aged 40 and above, the safe lower limit of blood oxygen is increased by 1% (from 93% to 94%) to ensure that the benchmark is in line with actual combat needs and age-related physiological characteristics.

[0076] Real-time monitoring and analysis subunit: Acquires trainees' heart rate and blood oxygen saturation data in real time, performs heart rate variability analysis and blood oxygen trend analysis, and outputs comprehensive load index based on the analysis results;

[0077] In this subunit, the trainee's heart rate and blood oxygen saturation data are acquired in real time, and the acquired data is preprocessed.

[0078] Heart rate variability was analyzed by extracting two core indicators: time domain and frequency domain.

[0079] Specifically, a fixed analysis window of 5 minutes is used, and the time interval between two consecutive heartbeats in the corresponding time period is extracted from the preprocessed heart rate data every 5 minutes as the heartbeat interval time.

[0080] If the number of valid heartbeat interval data points in a certain window is less than 240 (i.e., when the heart rate is greater than 120 bpm, there are only 300 heartbeats in 5 minutes, which may be insufficient after removing abnormalities), the window will be automatically extended to 10 minutes to ensure that the amount of data meets the analysis requirements.

[0081] Calculate the standard deviation of the heartbeat interval time within a fixed analysis window, calculate the difference between the standard deviation of the heartbeat interval time and the corresponding baseline reference value, and calculate the ratio of the absolute value of the difference to the corresponding baseline reference value as the standard deviation ratio.

[0082] Calculate the root mean square of the difference between adjacent heartbeat intervals within a fixed analysis window, which is the square root of the average of the squares of the differences between adjacent heartbeat intervals. Compare the root mean square of the difference between adjacent heartbeat intervals with the corresponding baseline reference value, and calculate the ratio of the absolute value of the difference to the corresponding baseline reference value as the root mean square deviation ratio.

[0083] By using Fast Fourier Transform, the heart rate interval data is transformed from the time domain to the frequency domain, and divided into low-frequency band (0.04-0.15Hz) and high-frequency band (0.15-0.4Hz).

[0084] Calculate the power values ​​of the low-frequency band and the power values ​​of the high-frequency band, and then calculate the ratio of the power values ​​of the low-frequency band to the power values ​​of the high-frequency band as the power ratio. Calculate the difference between the power ratio and the corresponding basic reference value, and calculate the proportion of the absolute value of the difference to the corresponding basic reference value as the power deviation ratio.

[0085] The heart rate variability analysis index is obtained by weighted and fused calculation of the standard deviation ratio, the root mean square deviation ratio, and the power deviation ratio.

[0086] It should be explained that the standard deviation of the heart rate interval reflects the overall activity of the autonomic nervous system and is the most core broad-spectrum indicator of physiological stress. It is sensitive to vestibular stimulation, and the standard deviation deviation ratio weight is set to 0.35. The root mean square difference between adjacent heart rate intervals reflects vagal tone and directly reflects the degree of vagal inhibition caused by stress. The root mean square difference deviation ratio weight is set to 0.3. The power ratio reflects the balance between the sympathetic and vagal nerves and accurately judges the level of sympathetic nerve excitation (a direct indicator of increased load). The power deviation ratio weight is set to 0.35.

[0087] Blood oxygenation trend analysis is performed using the blood oxygenation slope.

[0088] Synchronize with the heart rate variability analysis at a fixed analysis window (5 minutes). Extract blood oxygen saturation data for the corresponding time period from the preprocessed blood oxygen data every 5 minutes and record the timestamp of each data point.

[0089] Using the least squares method, a linear fit was performed on the blood oxygen saturation within a fixed analysis window to obtain the trend equation for blood oxygen saturation: Where y represents blood oxygen saturation, t is time, k is the slope, and b is the intercept;

[0090] The calculated k value is the blood oxygen slope;

[0091] The blood oxygenation slope is calculated by comparing it with the corresponding baseline reference value. The absolute value of the difference is then calculated as a percentage of the baseline reference value to obtain the blood oxygenation trend index.

[0092] The heart rate variability analysis index and the blood oxygen trend index are weighted and fused to output a comprehensive load index;

[0093] Among them, the comprehensive load index needs to integrate the scores of heart rate variability and blood oxygen trend, and set the weights based on the priority of their impact on physiological load (in anti-vertigo training, the weight of autonomic nerve stress is higher than that of respiratory and circulatory changes). The weight allocation is as follows: the weight of the heart rate variability analysis index is 0.7, and the weight of the blood oxygen trend index is 0.3.

[0094] Multi-level intervention triggering and execution sub-unit: Based on the comprehensive load index output, trigger tiered early warning and intervention execution;

[0095] In this subunit, the comprehensive load index is obtained, and the safety threshold for the corresponding trainee is constructed by combining the aforementioned benchmark: the comprehensive load index range corresponding to the three levels of intervention threshold is determined by combining the branch of service (e.g., pilot vs. ordinary infantry), age (≥40 years old vs. <40 years old), and individual historical data (e.g., the average index value triggered by intervention in past training).

[0096] If the timeframe falls within the Level 1 intervention threshold range and the duration is greater than or equal to 3 seconds, then Level 1 intervention is initiated, including:

[0097] Send a command to the VR immersive interaction module: a yellow flashing warning icon (such as the text "heart rate stress increase" + heartbeat icon) pops up in the lower right corner of the VR scene. The icon is displayed for 10 seconds and does not obstruct the core field of view of the training scene (such as tactical task test questions) to avoid interfering with cognitive tasks.

[0098] Push a Level 1 intervention prompt to the instructor console: Highlight the trainee in yellow in the trainee list and display the current comprehensive load index and triggering reason (such as HRV deviation ratio increasing to 35%).

[0099] Upgrade Decision: If the comprehensive load index does not decrease within 10 seconds (still ≥ the lower limit of the first-level threshold), or rises to the second-level threshold range, the second-level intervention will be automatically triggered.

[0100] If the timeframe falls within the secondary intervention threshold range and the duration is greater than or equal to 3 seconds, then secondary intervention is initiated, including:

[0101] Send a command to the headphone unit of the VR immersive interaction module: Play an offline voice prompt (e.g., if your physiological load is close to the safe limit, please keep your breathing steady and reduce violent head movements). The voice prompt is 3 seconds long and repeats once every 5 seconds. The volume should be 10dB higher than the ambient sound (to ensure that the trainee can hear it clearly).

[0102] Send a Level 2 intervention audio-visual alarm to the console: The console indicator light flashes yellow and a buzzer sounds (1 time / second), while an intervention suggestion pops up (such as suggesting to observe trainee A's gait and manually reduce the platform speed if necessary).

[0103] Send a pre-deceleration command to the three-degree-of-freedom electric rotary platform: the platform will reduce its current speed by 10% (e.g., from 12 rpm to 10.8 rpm) to buffer against possible subsequent level three intervention;

[0104] Upgrade Decision: If the comprehensive load index does not decrease within 15 seconds (still ≥ the lower limit of the second-level threshold), or rises to the third-level threshold range, the third-level intervention will be automatically triggered.

[0105] If the threshold range for Level 3 intervention is met and the duration is greater than or equal to 3 seconds, then Level 3 intervention is applied, including:

[0106] Send a smooth deceleration command to the three-degree-of-freedom electric rotary platform: the platform stops moving according to a linear deceleration curve (e.g., the speed drops from the current speed to 0 within 15 seconds, with an acceleration ≤0.2m / s²), avoiding sudden stops that could cause impact injuries to the trainee's lumbar and cervical spine; at the same time, trigger the platform's emergency stop indicator light (solid red) to alert those around.

[0107] Send a safety pause screen to the VR headset unit: The VR scene switches to a blue static interface, displaying the text "Training is temporarily paused. You have safely stopped the training. Please keep your sitting posture stable. At the same time, the cognitive task test questions will stop being pushed."

[0108] Send a high-frequency monitoring command to the physiological information acquisition module: increase the wristband's sampling frequency from 1Hz to 2Hz, focusing on monitoring the recovery trend of heart rate and blood oxygen;

[0109] Synchronous alarm to medical point / instructor: Send emergency support prompt (including trainee's name, location, and current physiological indicators) to the medical point in the training field through the communication interface unit. At the same time, a three-level intervention confirmation box pops up on the console, prompting the instructor to go to the scene to check.

[0110] Safety backup: If the platform is unable to receive deceleration commands due to mechanical failure (such as motor communication interruption), the system will automatically trigger the platform's physical power-off switch (a backup circuit independent of motor control) to forcibly cut off the power supply and ensure that the platform stops rotating within 10 seconds.

[0111] This embodiment has at least the following effects:

[0112] First, the benchmark construction sub-unit establishes initial benchmarks for both new and experienced trainees through resting monitoring and historical data extraction. Then, it combines a 10-minute sliding window algorithm for iterative calibration and makes personalized adjustments based on branch of service (such as the difference in the safe upper limit of heart rate between pilots and ordinary infantry) and age (the blood oxygen threshold is adjusted upward for those ≥40 years old). At the same time, it converts physiological indicators of different scenarios into relative resting value fluctuation percentages to eliminate scenario intensity interference. This ensures that the evaluation benchmark is both in line with individual physiological patterns and matches actual combat needs, significantly improving the scientific nature of training.

[0113] Secondly, the real-time monitoring and analysis subunit constructs a quantitative system through multi-dimensional indicators, transforming abstract physiological states into accurately quantifiable values, replacing traditional subjective methods such as gait observation and self-reporting, and achieving an objective and refined assessment of physiological load.

[0114] Furthermore, the multi-level intervention triggering and execution sub-unit matches personalized thresholds based on comprehensive load indicators to carry out multi-level early warning interventions, which not only ensures the safety of trainees (such as avoiding cervical spine injuries caused by sudden stops and preventing the risk of hypoxia), but also avoids blindly interrupting training, thus balancing safety and training continuity.

[0115] Overall, the closed-loop design of personalized benchmarks, quantitative analysis, and intelligent intervention not only optimizes the limitations of traditional anti-vertigo training equipment, but also provides support for subsequent training program optimization and big data analysis through data recording. This allows anti-vertigo training to shift from experience-driven to data-driven, improving the accuracy and safety of training. Furthermore, the design tailored to the actual combat needs of various branches of the armed forces (such as high-stress benchmarks for pilots and indicator assessments in tactical scenarios) strengthens the integration of training and combat, providing a scientific and reliable technical solution for modern military anti-vertigo training.

[0116] Example 3

[0117] During training, when two or more trainees trigger physiological emergencies simultaneously, due to the limited manpower of a single instructor, it is impossible to treat all trainees in an emergency at the same time. Only one person can be intervened first, while the other or more are left in a state of waiting for intervention. This may cause the physiological burden of those who are not intervened in time to continue to deteriorate, from mild stress (such as high heart rate) to serious emergencies (such as fainting, vomiting obstructing the airway, acute damage to vestibular function), missing the best intervention window. When emergencies occur in a concentrated manner, instructors may experience decision-making delays or oversights in intervention operations due to being overwhelmed, which may also affect the stability of the status of other normal trainees.

[0118] Based on the foregoing embodiments, please refer to Figure 4 As shown in the embodiment of the present invention, an intelligent military physical fitness anti-vertigo training device includes:

[0119] When multiple trainees are training at the same time, if two or more trainees are detected to have triggered the intervention condition at the same time, the multi-objective emergency priority analysis process will be triggered.

[0120] The target emergency priority analysis process includes:

[0121] Based on the real-time monitoring and analysis subunit in the aforementioned embodiments, the comprehensive load index of all emergency trainees is obtained.

[0122] Those emergency trainees at the same intervention level are grouped together as emergency trainees at the same level.

[0123] The reason for statistically classifying emergency trainees at the same intervention level is as follows: First, it avoids confusion in priority determination caused by mixing trainees at different intervention levels (such as Level 3 high risk and Level 1 warning). The same intervention level means that the trainees' comprehensive load indicators are in a similar safety risk range. After statistical aggregation, it can provide clear grouping for subsequent detailed ranking within the same level, and also allow the system to perform basic intervention actions at the same level in batches (such as simultaneously starting voice reminders for all trainees in the Level 2 intervention set), reducing redundant judgment operations; at the same time, it helps instructors quickly grasp the emergency scale of each risk level and avoid the imbalance of energy allocation caused by the dispersion of goals.

[0124] Based on a multi-level analysis strategy, emergency priorities are assessed, specifically as follows:

[0125] The comprehensive load indexes of emergency trainees of the same level are sorted in descending order from largest to smallest to obtain a preliminary priority ranking.

[0126] Calculate the difference between the first and second comprehensive load indicators of two adjacent data points in the preliminary priority ranking to obtain the indicator deviation value;

[0127] The index deviation value is compared with the preset deviation limit, which is used to distinguish two trainees with similar comprehensive load indexes. It is set by those skilled in the art based on experience and is generally set to 3%.

[0128] Two trainees whose indicator deviation values ​​are less than the preset deviation limit are identified and marked as trainees with similar indicators.

[0129] Two trainees whose indicator deviation values ​​are greater than or equal to the preset deviation limit are identified and marked as trainees whose indicators are not close to the target.

[0130] For trainees whose indicators are close to those in the trainee priority ranking, those with a health label are considered.

[0131] For example, trainee A with a history of vestibular injury is given higher priority than trainee B who is not labeled.

[0132] If the health indicators are consistent, sort them according to the combat needs of the military branch;

[0133] For example, pilot trainees are given priority over regular infantrymen because the former experience higher vestibular loads in actual combat.

[0134] If the combat requirements of different military branches are also consistent, then retrieve and calculate the change in the trainee's comprehensive load index within 10 seconds.

[0135] The changes in the comprehensive load index of each trainee are compared, and the trainee with the largest change in comprehensive load index has a higher training priority than the trainee with the smallest change in comprehensive load index.

[0136] If the changes in the overall load indicators remain consistent, other indicators should be selected for comparison, including but not limited to: blood oxygen trend indicators and heart rate variability indicators.

[0137] For example, if the four trainees are ranked in descending order of their comprehensive load index: Trainee 1, Trainee 2, Trainee 4, Trainee 3;

[0138] And through calculation, it was found that the indicators of trainee 1 and trainee 2 were similar, and the indicators of trainee 2 and trainee 4 were similar.

[0139] The health status check revealed that Trainee 1 had a "previous vestibular function impairment" status, while Trainee 2 had no health status. Therefore, Trainee 1 has a higher priority than Trainee 2.

[0140] If Trainee 4 and Trainee 2 have the same health status and neither has any underlying diseases, then they will be ranked according to "combat requirements of the branch of service": Trainee 4 is a pilot and Trainee 2 is a regular infantryman, so Trainee 4 has higher priority than Trainee 2.

[0141] This example has at least the following effects:

[0142] The system optimizes the safety hazards of simultaneous emergency response by multiple trainees, avoiding the risk escalation caused by low-priority waiting. Through "same-level classification + multi-level sorting", emergency trainees are first divided into groups according to intervention level, and then the priority is refined through rules such as comprehensive load index deviation value, health label, and military branch requirements. This mechanism ensures that high-risk targets are dealt with as soon as possible, avoiding a passive situation of neglecting one aspect while dealing with another.

[0143] This significantly improves the scientific rigor and objectivity of emergency priority determination, avoiding the uncertainty of human experience while ensuring that each step of the ranking is based on traceable data (such as prioritizing pilots over ordinary infantry to meet actual combat vestibular load requirements, and prioritizing those whose indicators rise sharply within 10 seconds to match dynamic changes in physiological risk), allowing multi-target emergency response to shift from experience-driven to data-driven.

[0144] Example 4

[0145] Based on the foregoing embodiments, such as Figure 5 As shown, this application provides an intelligent military physical fitness anti-vertigo training method, which specifically includes the following steps:

[0146] Step 1: Trainees log in through the facial recognition system, which automatically retrieves their personal profiles. Trainees then put on their biometric wristband, VR headset, headphones, and earpiece.

[0147] Step 2: The instructor selects the standard training mode (such as ship crew mode or pilot mode) or customizes the rotation parameters and question bank for this training through the equipment control interface;

[0148] Step 3: Start training. The three-degree-of-freedom rotation platform begins to move, and the VR headset scene starts synchronously. While enduring rotational stimulation, the trainee needs to observe the situation in the VR headset environment and answer the questions raised by the system.

[0149] Step 4: Collect and analyze physiological information data in real time. If any abnormalities are found, immediately implement intervention according to the preset level.

[0150] Step 5: After training, generate a detailed evaluation report, pointing out the strengths and weaknesses, and recommending the intensity and focus of the next training session based on historical data.

[0151] As a specific embodiment of the present invention: taking the training of a ship's crew member as an example;

[0152] The trainee stands in the center of the platform, facing the three-degree-of-freedom rotating platform. The identity is verified by facial recognition through a camera. The instructor selects the "moderate wind and waves" mode from the program library on the platform control system user interface. This mode is preset with a random wave motion curve that is mainly based on the roll degree of freedom and combined with a small yaw.

[0153] After the trainee puts on the wristband and VR headset, the platform moves according to the preset parameters. Zhang San's VR view shows the interior environment of the ship's cabin, which shakes synchronously with the platform's movement. At the same time, a question appears in the VR view: "Current heading 030, suspicious target detected at bearing 120, requesting our ship to turn to a new heading?" The trainee, while in a dizzy state, keeps thinking and answers through the headset: "Turn to the new heading 120 degrees." The system recognizes the voice and judges the answer to be correct, and adds points.

[0154] During the process, the trainee's heart rate was monitored to rise rapidly from 75 bpm to 135 bpm and remain at a high level. It was determined that the physiological load was too high, and a yellow warning icon immediately popped up in the VR field of vision (Level 1 intervention). After 5 seconds, the heart rate still did not decrease, and a voice warning was issued through the headphones: "Heart rate is too high, please keep breathing" (Level 2 intervention). After another 10 seconds, the heart rate continued to climb to 145 bpm, and the system automatically triggered Level 3 intervention. The control platform smoothly decelerated to a complete stop within 15 seconds.

[0155] After training, the system generated a report: Training duration 3 minutes and 45 seconds, average heart rate 128 bpm, maximum heart rate 145 bpm, correct answer rate 80%, average reaction time delay 2.5 seconds, overall score: 72 points (more points deducted in physiological stress category); system suggestion: "Next training, appropriately reduce the peak rotation intensity and focus on improving psychological stability under high stress conditions."

[0156] As a specific embodiment of the present invention:

[0157] After the user logs in using facial recognition, the AI ​​system recommends an initial intensity plan. Once training begins, the platform accelerates to the target speed in a sinusoidal curve. Simultaneously, a math problem is displayed in the VR environment: "15 + 28 = ?". The user answers "43" through the headset. The voice recognition system recognizes the answer and compares it with the input. If the answer is correct, a point is awarded. At the same time, the AI ​​evaluation submodule detects that the user's heart rate continuously exceeds a preset threshold (such as 85% of the age-related maximum heart rate). It then sends a command to the platform control submodule to decelerate the platform to a stop and displays a warning message "Training ended, please note" in the VR view. The training report ultimately displays the training duration, average heart rate, maximum heart rate, correct answer rate, and the comprehensive score given by the system.

[0158] The foregoing has shown and described the basic principles, main features, and advantages of the present invention. Those skilled in the art should understand that the present invention is not limited to the above embodiments. The embodiments and descriptions in the specification are merely illustrative of the principles of the invention. Various changes and modifications can be made to the invention without departing from its spirit and scope, and all such changes and modifications fall within the scope of the present invention as claimed. The scope of protection of the present invention is defined by the appended claims and their equivalents.

Claims

1. An intelligent military physical fitness anti-vertigo training device, characterized in that: It comprises a three-degree-of-freedom electric rotating platform, a physiological information acquisition module, a VR immersive interaction module, a platform terminal control and AI central system, and a database storage module. The platform terminal control and AI central module comprises a device control unit, an AI evaluation and intelligent intervention unit, an automatic scoring unit, and a communication interface unit. The AI evaluation and intelligent intervention unit comprises a benchmark construction subunit, a real-time monitoring and analysis subunit, and a multi-level intervention triggering and execution subunit. The benchmark construction subunit acquires the basic information and historical training data of the trainee, and sets the benchmark dynamically through a sliding window algorithm. The benchmark construction subunit specifically comprises: If the trainee has historical training data, the physiological indicators of the previous 3-5 anti-dizziness training sessions are extracted, and the initial physiological data closest to the current state in the historical training is extracted as the initial benchmark. If the trainee is training for the first time, a 5-minute resting state monitoring is guided, and the resting heart rate, basic blood oxygen saturation, and heart rate variability basic value collected are used as the initial benchmark. The collected historical data and resting state data are preprocessed by removing outliers, and the physiological indicators under different training scenarios after preprocessing are converted to a unified dimension. Taking 10 minutes of training duration as the sliding window unit, the initial benchmark is iteratively calibrated in combination with the data of the trainee's previous 3 low-intensity adaptive training sessions, and the benchmark is finally corrected in combination with the characteristics of the military branch. The real-time monitoring and analysis subunit acquires the heart rate and blood oxygen saturation data of the trainee in real time, and performs heart rate variability analysis and blood oxygen trend analysis, and outputs a comprehensive load index based on the analysis results. The multi-level intervention triggering and execution subunit triggers graded early warning and intervention execution based on the output comprehensive load index. In the multi-level intervention triggering and execution subunit, when multiple trainees are training simultaneously, if two or more trainees are simultaneously monitored to trigger the intervention condition, a multi-target emergency priority analysis process is triggered. The multi-target emergency priority analysis process comprises: Statistical emergency trainees at the same intervention level are taken as a same-level emergency trainee set. The comprehensive load index corresponding to the same-level emergency trainee set is sorted in descending order from large to small to obtain a preliminary priority ranking. The difference between the first and second comprehensive load indices in the preliminary priority ranking is calculated to obtain an index deviation value. Two trainees with an index deviation value less than a preset deviation limit value are extracted and marked as index proximity trainees. The index proximity trainees are prioritized according to the health annotations of the trainees. If the health annotations are consistent, the trainees are sorted according to the needs of the military branch. If the needs of the military branch are also consistent, the comprehensive load index change amount of the trainee within 10 seconds is calculated. The comprehensive load index change amounts of the trainees are compared, and the trainee corresponding to the maximum comprehensive load index change amount has a higher priority than the trainee corresponding to the minimum comprehensive load index change amount.

2. The intelligent military physical anti-dizzy training device according to claim 1, characterized in that: The three-degree-of-freedom electric rotating platform has three degrees of freedom in pitch, roll, and yaw, which are driven by independent frequency-modulated servo motors. The physiological information acquisition module acquires the physiological indicators of the trainee in real time. The VR immersive interaction module comprises a VR head-mounted unit, a display screen unit, and a headset and earphone unit; The VR head-mounted unit is used for simulating a specific military movement scene; The display screen unit is used for superimposedly displaying test questions in the VR scene; The headset and earphone unit is used for collecting the voice answer of the trainee, and the earphone is used for playing environmental sound effects and instructions.

3. The intelligent military physical anti-dizzy training device according to claim 1, characterized in that: The process of the heart rate variability analysis is as follows: Through time domain and frequency domain analysis of the heart rate data, the standard deviation ratio, the difference root mean square deviation ratio, and the power deviation ratio are calculated; The standard deviation ratio, the difference root mean square deviation ratio, and the power deviation ratio are fused to obtain the heart rate variability analysis index.

4. The intelligent military physical anti-dizzy training device according to claim 3, characterized in that: The process of the time domain and frequency domain analysis is as follows: A 5-minute fixed analysis window is used, and the time interval of two consecutive heartbeats in the corresponding period is taken from the preprocessed heart rate data every 5 minutes as the heartbeat interval time; The standard deviation of the heartbeat interval time in the fixed analysis window is calculated, and the standard deviation of the heartbeat interval time is subtracted from the corresponding basic reference value to calculate the proportion of the absolute value of the difference to the corresponding basic reference value, which is taken as the standard deviation ratio; The difference root mean square of adjacent heartbeat intervals, i.e., the average value of the square of the difference between adjacent heartbeat intervals, is calculated, and the difference root mean square of adjacent heartbeat intervals is subtracted from the corresponding basic reference value to calculate the proportion of the absolute value of the difference to the corresponding basic reference value, which is taken as the difference root mean square deviation ratio; The heartbeat interval time data is converted from the time domain to the frequency domain through fast Fourier transform, and the low frequency band and the high frequency band are divided; The power values of the low frequency band and the high frequency band are calculated, and the ratio of the power values of the low frequency band and the high frequency band is calculated as the power ratio, which is subtracted from the corresponding basic reference value to calculate the proportion of the absolute value of the difference to the corresponding basic reference value, which is taken as the power deviation ratio.

5. The intelligent military physical anti-dizzy training device according to claim 1, characterized in that: The process of the blood oxygen trend analysis is as follows: The blood oxygen saturation data in the corresponding period is taken from the preprocessed blood oxygen data every 5 minutes, and the timestamp of each data point is recorded; The least square method is used to linearly fit the blood oxygen saturation in the fixed analysis window to obtain the trend equation of the blood oxygen saturation, and the blood oxygen slope is calculated; The blood oxygen slope is subtracted from the corresponding basic reference value to calculate the proportion of the absolute value of the difference to the corresponding basic reference value, and the blood oxygen trend index is obtained.

6. The intelligent military physical anti-dizzy training device according to claim 5, characterized in that: The process of outputting the comprehensive load index is as follows: The heart rate variability analysis index and the blood oxygen trend index are fused to output the comprehensive load index.

7. An anti-vertigo training method based on the device according to any one of claims 1 to 6, characterized in that, The following steps are included: The trainee logs in through the face recognition system, and the system automatically retrieves his / her personal archives, and the trainee wears the physiological bracelet, the VR head-mounted device, the earphone, and the headset; The instructor selects the standard training mode or customizes the rotation parameters and the test question bank for this training through the device control interface; The training is started, the three-degree-of-freedom rotation platform starts to move, the VR head-mounted scene is started synchronously, and the trainee observes the situation in the VR head-mounted environment and answers the questions raised by the system while bearing the rotation stimulation; Real-time physiological information data collection and analysis, if abnormal, immediately according to the preset level to execute intervention; After the training, generate a detailed evaluation report, and recommend the next training intensity and focus based on historical data.

Images